Optimal Solutions in Network Theory for Influencer Detection

Highly influential accounts can be detected simply.

There are many practical applications for doing so:

The optimization problem here is constructing a network that is fully connected at lowest cost.

Short-term divergence and micro-evolution in populations can be reconstructed based on sample data.

The same principles can be applied to parallel computation, social networks, transportation/supply design and power networks.

Some assumptions in modeling the system include:

There exists acritical massfollowing to trigger viewing or dispersal.

There existsorganicandsyntheticinteraction with followers to encourage likes and comments.

Influence reaches a specific or distinct community, unlike a celebrity which may have a more random distribution of followers.